How Should Education Companies Like New Oriental Respond to AI Digital Natives Like LAIX? (Tech Strategy – Podcast 35)

In this class, I go into the how digital native education companies like LAIX are challenging traditional companies like New Oriental.

You can listen here or at iTunes, Google Podcasts and Himalaya.

Which of the 6 Digital Superpowers does LAIX have? If any.

  1. Dramatically transform the user experience
  2. Enable a platform business model
  3. Network effects
  4. Other competitive advantage
  5. Virality
  6. Scalable

Related podcasts and articles:

This is part of Learning Goals: Level 2, with a focus on:
  • #1: Basics of Digital Economics

Concepts for this class:

  • Jobs to be Done
  • Digital Econoimcs
  • Digital Superpower 1: Dramatically Transform User Experience

Companies for this class:

  • Liulishuo (LAIX)
  • New Oriental Education


I write, speak and consult about how to win (and not lose) in digital strategy and transformation.

I am the founder of TechMoat Consulting, a boutique consulting firm that helps retailers, brands, and technology companies exploit digital change to grow faster, innovate better and build digital moats. Get in touch here.

My book series Moats and Marathons is one-of-a-kind framework for building and measuring competitive advantages in digital businesses.

Note: This content (articles, podcasts, website info) is not investment advice. The information and opinions from me and any guests may be incorrect. The numbers and information may be wrong. The views expressed may no longer be relevant or accurate. Investing is risky. Do your own research.

——Transcription below

Welcome, welcome everybody. My name is Jeff Towson and this is Tech Strategy. And the question for today, how should traditional education companies like New Oriental in China respond to AI digital natives like Lioli Shuo? So it’s this very traditional service-based, very local business of education. private in this case, being challenged by these digital first companies who are really relying on AI more and more. How do they respond? What should they do? Should they do anything? Is this a major threat? What would you do if you were advising these companies? What should they do? I think that’s a pretty good question for today. It’s a very common situation where you have a traditional well established business, especially service based. that is seeing this stuff more and more emerge into their world. And usually the people running those companies, this is a new subject for them. So that will be the case for today. And we’re gonna do this as a case. I’m gonna give you some background, give you some frameworks for thinking about it, and then ask you to take a stab at it. What would you advise these companies to do? But first, a welcome to the new subscribers we have this week. We’re doing well, things are moving forward, more and more people are signing up, which is great to see. And welcome to the class. If some of this stuff is new, most of it is probably new, and you feel a little overwhelmed, don’t worry about it. It’s sort of a step-by-step process. We’ll tackle ideas in companies one after the other after the other, repeat and repeat. ideas and concepts and ways of thinking about digital that were foreign to you and new to you will slowly become very common and easy and it’ll all become familiar. So it’s a gradual process. So welcome and for those of you who aren’t subscribers please go over to You can sign up there. There’s a free 30-day trial. See if you like it. And now let’s get into the case. Now first, let’s just sort of talk about education technology. I’ve been looking at education technology, EdTech, in China for four years now. And it’s been a big subject in theory for a long time, but not a big subject in practice. Because everyone would look at the China story and they’d say, oh my God, look at all these people studying English because you have 300 million plus people studying English and the factoid you always hear is, oh, there are more people studying English in China than speak English in the US. And big population, education is very, very important. The spending on education is fantastic. Like… Chinese families spend an enormous amount of money on private education, K through 12. I mean, it starts very early in life. You’re talking five-year-olds, four-year-olds are starting to learn languages and things like that. And then they go through high school. It’s very competitive. You got to do well on your tests. You want to pass your foreign language exams. You want to, everyone wants an edge. So the spending is huge and it’s mostly at the family level. So everyone has sort of looked at the China education market for a long, long time as a big potential opportunity because the spending’s already there in the offline traditional world. And the idea was, oh, this is gonna move online. It’s gonna happen on phones. We’re gonna see digital natives. We’re gonna see online tutoring, language, test prep, a lot of that stuff. And it never really happened. It wasn’t the big boom everyone thought it was gonna be. Similar thing happened in healthcare, which I’ll probably talk about at some point. It was a huge idea, there’s a lot of venture capital money and went into it in a couple years, very excited, a lot of hype, and then it just sort of quietly faded away. And that is a real contrast to say retail, where there was a lot of excitement and retail is changing very, very quickly. Communications, media, social media. All of those things are changing very, very fast. Transportation, moderately fast. But education and healthcare use even worse, really has been sort of slow. You know, first of all, you have to define what we’re talking about education, because that covers a lot of stuff. Generally, what we’re talking about is English language, which is a lot of people, adults and high schoolers. studying English in some form, in person, online, or whatnot, usually with the objective of becoming more and more proficient, getting their language skills up, and probably passing a foreign test such that then they can apply to school or whatever in UK or the US. And they can show the certificate, I’m a level so-and-so, I passed the test. So it’s a lot about proficiency. It’s a lot about test preparation. That’s a big bucket in the private education world of China. Another one, we’ll just broaden that out, we’ll say at the high school level, in particular, you have a huge amount of test prep. Everyone is competing for the Gaokao. Everyone’s trying to get ahead and score high. Everyone’s trying to do foreign exams, the MCAT, whatever. Business school, medical school, law school, everyone’s prepping for these various tests. So there’s a whole lot of services in preparation. A lot of that at the high school level, also at the college level, things like that. So there’s a lot of test prep going on. That’s another big bucket of private education in China. Another bucket would just be, I would call it the get ahead market, which is biology, chemistry, physics, pretty much every subject that you would learn in high school. People are probably taking private courses on Saturday, Sunday. to just get ahead and do better in their classes, everything’s competitive. So that really, English is one sort of narrow bucket of content, learn a language. The test prep is a larger bucket, past the LSAT, past the MCAT, past whatever. That’s a little bit larger bucket, but it’s also relatively self-contained. But when you move into just sort of the get ahead market, Then you’re talking about every subject, literature, Chinese literature, English literature, physics, math, chemistry, whatever, that’s a huge sprawling set of content. And that’s usually a lot of high school stuff. And then you can even go younger, where you see K through 12 elementary school kids are studying on Saturday, Sunday, everyone’s taking extra classes, things like that. So it goes all the way down. But that’s kind of, you hear that, that sort of framework discussed a lot when it turns out What are people spending money on family-wise to help their children or help themselves in education? A lot of it is those three to four buckets. And so when you look at education, one of the first things you have is you have the sort of Clayton Christians in question of what is the job to be done? And this is, I’m sorry, there’s three main concepts for today I should have said earlier that are part of the course. The first one is jobs to be done. Second one is the sexy but dangerous economics of digital. And number three is digital superpower number one, transform the user experience. So those are ideas I’ve already talked about previously. But yeah, the first one is really jobs to be done is what is the point of education? Like what is the job that it is fulfilling when you go to a local school down the street to teach your eight-year-old English? or you go to prepare, you know, take a test prep course, or you go to college, you know, other types of education beyond what I’ve just mentioned, what is the job you are hiring this company to do for you? And when you take that apart, you realize, particularly colleges are really doing multiple different jobs and graduate school really are doing a lot more. I would put it in sort of four buckets of jobs to be done. Number one, I think you’re learning an advanced skill. You’re learning something that you probably can’t learn on your own. You can learn a lot of stuff on your own. You can learn to throw a frisbee. Yeah, you might be able to learn some basic skills. You can learn card tricks. I mean, you can learn a lot of stuff online, reading books, you can learn to cook some stuff. But there’s certain advanced skills that you really do need a structure. to learn like a language, most people, almost everyone who learns a language has a teacher or goes to a school. This course, I mean, I’ve specifically focused this course on things I don’t think people can learn on their own, where you need a structure, you probably need a teacher of some kind. Sort of advanced skills would be job number one. Within the advanced skill, you could put it into a couple buckets, you could put it into situations where you’re just looking for basic proficiency. versus where you’re looking for excellence. Now, if you’re doing test prep, you’re not really looking just to be basically okay at mathematics such that when you take the test to get you into college or grad school, you’re proficient. No, you’re not looking for proficiency, you’re looking for excellence. You’re looking to be better than everybody else. Other skills, if you’re an adult and you’re just taking language proficiency on the side. Okay, you’re not looking to be in the top 2% of English speakers, you’re just trying to get, become more proficient, your baseline is higher. So you can sort of put those into like, are you trying to be the best, or are you just trying to get some basic skills going within this? And you can also think about skills, advanced skills that are static versus evolving. If you learn English or Chinese, it’s pretty much a static skill. Once you’ve learned it and you’re 80% solid at it, that’s pretty much it. It’s not like the whole language changes every three to five years, and everything you knew was obsolete, and now you have to learn something new. You know, this course is a lot about dynamic education, that things are constantly moving. We’re learning, we’re studying what’s happening this year. If you studied digital stuff from 2005, it’s not gonna help you very much right now. So within advanced skills, you can sort of break it down into those categories and others. Another job to be done, you could call it a credential. Look, maybe people aren’t going to get their MBA because they really want to learn accounting. Maybe they’re going to MBA to get the MBA. Maybe it’s about the credential, which you could see it as an advantage in your career. I have an MBA, that person doesn’t. You could see it as a type of career insurance by having a degree from a certain institution, undergrad or graduate, by having a graduate degree and an undergraduate degree. you get yourself some level of security in your career that you can always get a job. So you could call it maybe a type of credential, type of career insurance. That could be a second job, a second job to be done. The third one, you could say it’s about getting a job. This is why people go to MBA programs. You know, they’ll really go to MBA programs to learn accounting. No offense to anyone who teaches accounting, but it’s not that thrilling of a subject. It’s necessary, but… People go to MBA programs because they want the job placement when they’re done. They wanna go get their MBA from Wharton and then begin at Morgan Stanley the day after they graduate. And that’s really what matters. I mean, once you have that associate position, what comes before doesn’t really matter anymore. So it’s about a career transition. It’s about getting a job, things like that. Stepping stone in your career pathway, things like that. So you could say it’s a part of that. The fourth bucket you could put it in, it’s about the network. that education to some degree, it’s about who you meet. You meet your classmates who then go on in life to do other things and you know them and you all rise up in your industries together and suddenly your old friends are now CEOs. You could argue that education, especially from a top university or a top business school, is a, it’s a, what they call a slow network effect, that people come out of these schools and oh, they’re all Harvard students. Well, then they go and they join McKinsey and Goldman or Procter and Gamble or whatever company they want to go do venture capital, whatever. And then they rise up the ranks and then they become managers and they become the people that hire. And what do they look for when they hire? Well, they look for people that they think are good, which usually in practice means they look for people like themselves. So Harvard people tend to hire other Harvard people or other Ivy League people. And so therefore, the more people that have Ivy League degrees, the more valuable those degrees become to everybody involved in that network over time. So it’s kind of like a slow network effect that plays out over like 10 to 15 to 20 years. So you could say, look, there’s four different jobs to be done here. And some places like Harvard Business School, they’re probably a lot about the credential. They’re probably a lot about the network. They’re probably a lot about getting the job. And to some degree they’re about the skill, but you can learn accounting anywhere. Now, the categories we’ve just talked about, or I’ve just talked about, most people, why are they going to do English language? Well, if they’re an adult, they’re probably going for the skill, job number one. They’re probably not getting a job directly by attending. There’s no network effect. The credential probably helps a little bit if they pass their various credential, that helps them a little bit. It’s probably most about the proficiency. If you’re doing test prep, or you’re doing English language certification so that then you can apply to school in the US or Europe, you know, you could see that would matter. So that would be more about the credential, about getting a job, which would be getting into the school in that case. So you can see it would be a little bit different if you’re doing K through 12. just get ahead stuff where you’re studying biology and chemistry in after school programs in Beijing as a sophomore in high school, that’s probably mostly about the advanced skill, just getting smarter about the subject so that you are better positioned in life. And usually that’s, you know, it could be a basic skill or a more advanced skill. Okay, so let’s go back to this question of what is it that Chinese parents are paying for when they’re sending their children K through 12? to all of this private education, which they’re absolutely doing. And a lot of it is those first two buckets I just mentioned. A lot of it’s English, a lot of it’s test preparation, whether it’s to pass the gaokao and get into college in China, or get into programs, graduate schools, whatever. And then outside of that it’s the get ahead thing. Everyone’s just trying to get better skills, get better biology, get better mathematics. It’s just, it is just culturally Everybody does it. You know, anywhere you live in China, go down any residential street, and within two blocks, you’ll find a couple things. You’ll find a 7-Eleven type store, convenience. You’ll probably find a market. And then you will find some sort of local education center. And you can always find them because you just look out your window and you look on Saturday morning where all the young people are going. and they all walk into the local testing or local learning center down the street on every random street, and they’re studying there on Saturday. I mean, it’s just, everybody does it. And in terms of jobs to be done, they’re doing test prep, they’re doing English language, and they’re doing get ahead, lots of subjects. But for concept number one for today, jobs to be done, we would probably put all of that under this sort of advanced skill. idea. Maybe a bit about the credential or career insurance or get a job, but mostly I think it’s just about building up the skills because they’re very competitive. I think that’s most of what’s going on and that is very different than the reason people go to college or they go to business school. So it’s a simpler version of what we would see in an MBA program. And the company we can sort of think about in this area is New Oriental, which is, I don’t think it’s important to know this company, I’m just using it as an example. You know, it’s a publicly traded Chinese company. It’s called New Oriental Education and Technology Group. It’s Dongfeng, for those of you in China. You know, it’s one of the early movers of private educational services in China. They have centers everywhere. I’m looking at the Wikipedia thing right here, which is not the best source of information, but I think it’s fine for this case. As of 2016, New Oriental has 67 short time. language educational schools, 20 bookstores, 771 learning centers, and more than 5,000 third-party bookstores in 56 cities across China. They’ve had over 26 million students enroll. On an ongoing basis, they probably have about one to 1.5 million people enrolled in their various schools anywhere in China. Most of their education is coming through K through 12 tutoring services. elementary, middle, high school, which has a lot to do with test prep, it has a lot to do with getting ahead in life, things like that. And this sort of traditional business model for education, which is really local services. You open a center, you hire a local teacher, everyone walks down on a Saturday, everyone sits in class. This is the idea, this business model that everyone thought was ripe for digital disruption for many, many years. And the argument is pretty powerful. because it’s a really stupid business model. I mean, this is literally the exact same business model that we had 5,000 years ago. It hasn’t evolved. Everybody goes into a room, everybody sits down, one teacher stands at the front of the room and speaks. It’s a one-size-fits-all education. You know, this traditional, and if we want to open, we want to do twice as many students, we got to open twice as many centers. And if we have a center in Beijing that doesn’t do anything for us in Shanghai, we have to open up another service there. Local services, you have to keep opening and opening and opening and opening. And it’s a teacher plus a room. Uh, there’s a lot of problems with it. I mean, on every level it’s bad. First of all, it doesn’t scale at all, right? If you have one center in Beijing that doesn’t help you in Tianjin, you have to open another one. Then you have to open another one. That’s why these companies like New Oriental took decades to grow to the size they are. The quality is inconsistent. You might have a good teacher teaching this class, but your teacher in that class might not be very, very good. There’s no personalization. So if you have a good teacher teaching in one class with 30 students, it doesn’t mean all 30 of them are learning the same. One student could be picking up all of it. One could be totally lost. So, the quality in terms of the teachers can vary, but the quality in what the students are learning can vary. And also it’s not personalized. Like the fact that everyone is getting the exact same content makes no sense. If I was teaching, let’s say vocabulary in Japanese or Chinese or something, and it was 20 words for this lesson, some students might know 10 of them, but not the other 10, but another student… might know these 15, but not these five. Everyone would have different levels of understanding of those 20, yet I’m telling the whole class the exact same thing. What I really wanna be doing is tailoring the content for that day’s class to address the gaps of each individual student. You wanna tailor content the same way YouTube personalizes what you see online. It shows you what you specifically want to see. It’s a very personalized one-to-one media. Well, education should be one to one. It should look at what you understand and what you don’t understand of the things we’re dealing with, and then give you specific content to address what you need, not just broadcasting the same thing to everybody as everyone sits in the room and writes stuff down. So quality varies on the teacher level, quality varies on what people are learning. There’s no personalization, it’s not scalable. It’s fairly high cost. because you have to rent a space, you have to hire a teacher, which requires a certain amount of training, especially if you’re teaching something more advanced like physics. Very hard to open a thousand centers across China that teach physics, because you have to get people who understand physics. It’s easier if you’re doing English. And then it’s very inconvenient. You have to show up at this time of day on a Saturday, sit down for an hour. You know, it’s the anti-Netflix model. Netflix is, hey, you can watch anything you want to watch whenever you want to watch, wherever you are. 2 a.m., you can open up your iPad and watch whatever show you want, no commercials, all of that. So it’s very inconvenient. The time and the location are fixed. So now this was common in businesses in 1990, but a lot of businesses that had this model have all been wiped out. And that brings us to a Liu Lishuo. If New Oriental is sort of your prototypical traditional education business, local service business model, Liu Lishuo, which is L-A-I-X, people have a hard time pronouncing the Chinese name, so they usually just say L-A-I-X. It’s traded in New York under L-A-I-X as the stock code. Okay, they are the other extreme. They are an AI first company with no classrooms, no teachers. They’re not just trying to digitize education. They’re trying to replace teachers almost entirely with AI. So that’s the other extreme. So I’m giving you sort of two extremes and how should a company like New Oriental respond to this sort of digital native AI first company that is all about education. So Lio Lishuo. founded 2013 in Shanghai by a young man named Wang Yi, who is, I’ve actually met him at a conference a couple of years ago, I heard him speak. He’s a kind of an interesting guy, and his whole thing was we’re gonna, he was an AI person by background. We’re gonna focus on English learning products that are based on AI. So within these various types of education, I mean, he was focused on, we’re gonna teach people English, that’s what we’re gonna do. So very specific niche. and we’re going to be AI first. We’re not going to have classrooms. We’re not going to open centers anywhere. It’s all going to be on your smartphone anywhere you want. Now, how they describe themselves in their, you know, the last 20F I looked at was AI powered English learning products and services for individual users and corporate customers. Okay, English learning products, basically with AI. personalized learning, this is their comments, personalized learning by application of deep learning and adaptive learning technology. So right off the bat, this is like the first line of their 20F. I mean, right off the bat, they’re talking about personalization, that this one to many, one size fits all education model is a major weakness. So we’re gonna go personalized each person to help people improve English language skills anytime and anywhere. That’s their language again. So, They’re talking right off the bat, I would put this in, look, they’re definitely targeting convenience and personalization as big levers, which are really ridiculous weaknesses of traditional education. Now the core tool they are building, I mean, they’re a software company, so you got people building something every day. What they’re building is they call the AI teacher, and it does three functions. This is from their latest description, is it does hearing, their words. which is basically speech recognition and scoring. So that’s, you speak into your phone in English because it’s an English proficiency app, and the AI converts that to text. That would be one skill. That would be speech recognition and actual language processing. And then it scores your text on how well you spoke, how proficient are you, how good was your grammar, how good was your pronunciation, things like that. So that’s the second thing they do is evaluation. They check your pronunciation, they check your vocabulary, they check your fluency, they check your coherence. And then the third thing their AI teacher does is feedback. It gives you real time and personalized suggestions. Here’s the next article you should read. Here’s five exercises you should do based on what you just said into the phone. And our assessment of your pronunciation, your scoring, all of that. Here is some real time feedback that makes it a personalized teacher, basically. And the argument is what they’re building is a big, huge database and some very sophisticated algorithms in terms of assessing English as spoken by Chinese. So it’s not assessing perfectly spoken English, it’s assessing people who have certain different levels of ability, certain levels of pronunciation, understanding, you know, it can assess people’s basically. people who are native Chinese speakers trying to speak English, and it can assess their pronunciation and proficiency levels. That’s kind of the core engine of what they’re building. And they offer several services. These are all mostly freemium model, which is a good business model. One of the nice things we talked about before was the economics of digital, and it allows you to give things away for free, which is a good tool. So their flagship app, Basically it’s free and paid courses. You can do a lot of stuff for free. Try courses, speak into it. It’ll assess your English and it’ll check how you’re doing, your pronunciation, your grammar, things like that, and it’ll give you recommendations for the next exercises, things to do. And then they have paid apps, which are, you know, you step up to a freemium. So Dongni English course, which is personalized English proficiency levels. It systematically moves you up over time. As you do it longer and longer, it assesses your progress and it… It helps you progress over time. There’s eight levels they offer. They also have some sort of, they have a bit of an offline component. The main thing is the AI engine, but then they have an offline component, which is basically study groups that happen through WeChat. So you then get into private groups where you can talk to other people and they have what they call online study advisors. So these are people who are not teachers, but they’re sort of. customer service like staff that you can then go into these private groups and chat with them and that’s part of the premium service they have about a thousand plus of them Then they have another one don’t need select which is even more premium services Where you can get human English teachers as well, so they start to incorporate a human element into this Based on what level you’re paying for but the core thing is AI driven And then they have another app called the Lowly Shuo Reading App. And if you look at their latest financials, they’re showing you know, rapid revenue growth. I won’t go through the numbers, but basically it’s going up pretty quickly. However, their marketing and sales expense is also going up at about the exact same level. In fact, year after year their marketing, sales and marketing expense is roughly equivalent to their total revenue. So it’s that… scenario which is usually pretty worrisome. If this was a media company would be a huge red flag. But they do say they have 10 million monthly active users now up from about nine or eight a year ago up from four or five the year before that. So they say that’s increasing in about three million paying customers. So okay, I mean if they have three million people paying you know that’s real. I think the big marketing spend is concerning and also even though they say they’re an AI company, which is their core skill, you know, of their, let’s say 1600 employees, 1100 of them, 1200 of them work in sales and marketing. That’s not what you’d expect to see in an AI first company. You’d expect to see the vast majority of people doing coding and R&D and very little sales staff. So, you know, they look, even though they say they’re AI and digital first, Their staffing looks a lot more like a traditional education company that’s doing a lot of knocking on doors and selling person to person. But we’ll see. I’m not totally sure what all those sales and marketing people are doing, but it definitely got my attention. So compare that business model to the one I just described for New Oriental. I mean, New Oriental was like local services, open centers, hire people, one size fits all. You have to show up at this particular time and location to be caught. non-scalable. And then that’s the other extreme is is low leash, which is highly personalized, where, you know, every single person has their own unique specific course designed, and constantly being redesigned to help them advance level after level in their pronunciation, their proficiency, their grammar, their vocabulary. And it’s a lot of real time feedback. That’s, I think, a big lever in terms of an offering a product. The other thing that jumps out about Liao Lushuo is their cost structure is dramatically lower. I mean, ridiculously lower. It’s like when bookstores got wiped out by Amazon. You have a business model that doesn’t require you to open 770 learning centers across China because it is time and space independent. You know, they can, everyone in China, the day this thing goes online, everyone in China can access it from wherever they are. So no staffing costs in terms of teachers in every little city and every little street, no real estate costs, no desks, no furniture, no physical books. I mean, it is a very, very low cost model. Super convenient, right? It just wipes out this idea of… You know, our class, here’s our class schedule. It’s 9 a.m. on Saturday. Now, you can sit up at 2 a.m. anywhere you are and do your classes whenever you feel like it, and you can do it on the train, and you can do it on your phone, and you can do it on your laptop, and anywhere you want, anytime you want. You can do 100 a day, like binge watch it, or you can do whenever you feel like it. Total convenience, completely scalable, right? So, that’s a big contrast to traditional services. So there’s two models there. And in between those two, I think there’s another model, which we call sort of a halfway point. The problem, I think, with Leo Lichuo is it’s not clear that teaching people with software actually works. If you had an online physics class and Let’s say you were doing your online physics class that was just an app, it was all AI, it had a beautiful business model, totally scalable, because you’re studying for the GAUCAU, because you wanna get a really high score and get into a good college. Are you gonna do the online version that’s all AI? Are you gonna go down to the local street where there’s a professor or a teacher who teaches physics? And that’s the one everyone’s been going to for 20 years. And you could see at a certain point you’re like, I need the one that works. I can’t mess around here. I can’t try a risk on something that’s never been tried. And it may not work as well. Maybe it works okay, but maybe the teaching in person works better. So there is a question of, you know, how good is this? Can you actually learn a language well with just an AI-driven model, with an AI teacher and not a real teacher involved? I think it depends what your job to be done is. In some cases, it’s probably fine. I could do an AI-driven Chinese class online as an adult who’s just looking to moderately improve my proficiency in pronunciation. That would be fine with me. I wouldn’t do this if I was studying to get into an MBA program and I really needed to nail the GMAT. You know, I’d want a proven model with a teacher because I’m not willing to take a chance. Right, so you can see. Depending on the job to be done and the type of content, this may or may not work. And there’s also the question of does it even work? Do students who do this actually improve? You know if you sign up for a local class in Japanese and you go every weekend in two to three years, you are gonna speak Japanese well. You know that almost for sure. If you spent two to three years doing this, could you end up not being very good? Maybe. So I think there’s an effectiveness question on this. purely digital AI model. And I think there’s a perception and confidence question, especially by parents who are probably the decision makers on this. Okay, in between that model and the Dongfang New Oriental, you have sort of a halfway version, which is like a hybrid where, okay, we have, let’s say, online one-to-one teaching. It’s a company called 51 Talk, which I’m not gonna go into, but… You can basically go on your phone and you can connect through a platform business model with an English teacher in the Philippines, if you’re in China. And you can have one-to-one lessons with a credentialed English teacher who speaks fluent English and just do it on your phone. And it’s one-to-one, so there’s a real person. So you’re not saving on the labor costs, but you don’t have to open a center in every street in Beijing. You don’t have to train people locally and it is personalized, it’s one-to-one. So that’s not super low cost, but it is lower cost and it is personalized. So one-to-one online English teaching, which is a common app people use like 51 Talk. That’s sort of a halfway model between Liu Lishuo and say New Oriental. It’s not super low cost, but it is lower because we don’t have the centers, it’s personalized. It doesn’t have geographic constraints. Anyone sitting in a rural town in China can access that, which they probably couldn’t access in-person education because you’d need a local center, which is probably not going to be in a mountain village. And it has also the convenient time and location because you can arrange with your teacher when you want to do your class on your time. So it’s definitely hitting a lot of those buttons, but it’s not hitting all of them like Lui Lishuo. But people are also more confident that it will work. All right, so here’s the question for today. Now let’s say you get a call from New Oriental, they hear you’re good at digital thinking, and they say, come on down and tell us about this new company, Lodlish, well, and should we be concerned about this? Is it a major problem? Is it a minor problem? Should we hit the panic button? Should we just wait and see? Do we have time to build our own? Do we need to build our own version of it? Should we call up some bankers and look for something to… You know, how big of a threat, how big of a problem is this? And so you go down there, you show up, and I’ve sort of, my little cheat sheet I’ve given you is my six digital superpowers. This is just a rough, crude assessment. Does this company have one or more of the six digital superpowers? Because if it does, it is potentially a big threat. the six of them, I’ve written them down in the show notes, what the six are, but it’s basically, digital superpower number one, it dramatically transforms the user experience such that the previous offering becomes obsolete. When you use WhatsApp or Skype or WeChat, depending where you are, are line. The idea of using an SMS message on your phone from your mobile carrier becomes ridiculous because it’s free, right? They just dramatically transformed the user experience. Everyone uses WhatsApp, WeChat, Line, whatever, to send text messages. Virtually nobody ever uses that old, you know, text message you get from your mobile carrier. Netflix, same thing, like, you know, you can watch whatever you want anytime you want. in a huge bundle that costs you $5 per month, as opposed to having a cable package that costs $120 per month if you’re in the US. And it’s fixed times that you have to watch each show and you can only watch on your TV at home, much less convenient, much more expensive. So that’s always question number one, does it dramatically transform the user experience? And in this case, the user, actually you should think about who the user is. Is it? Is it the student or is it the parent doing the buying? That’s a little bit different. Okay, digital superpower number two. Does this enable a platform business model? Because if a platform business model can be created, an Uber, an Alibaba, an Airbnb, they have a lot of strengths on their own. And that tends to happen when a digital, when a certain industry that’s traditional starts becoming more digital. Often you will see a platform business model emerge and be fairly devastating. Number three, does it have a network effect? Again, those are usually with platform business models, but not always. Number four, does it have some other type of powerful competitive advantage, switching costs, share of the consumer mind, things like that. Number five, is there a viral mechanism? Group buying, social commerce, things like that. payment platforms. And number six, is it scalable? Now that’s my standard six digital superpowers. For those of you who are new, don’t worry about if you didn’t understand those. I’ve touched on all of them in previous lectures. Don’t worry about it, I’ll keep repeating them. They’re listed in the show notes. When you go in to meet with the CEO of New Oriental, they ask you, you know, I want you to answer for them, this company does or doesn’t have. digital superpower or more than one and here’s why. So that’s kind of the question. So pause the recording, take a look at the list and decide for yourself is Liu Lishuo does it have these or is it trying to build one of them? Okay so do that now, hit pause, look at the list, it’s in the show notes, decide if this has one or more digital superpowers in play. Pause now. Okay, how did you do? I mean, one of the reasons I give you these frameworks is when you’re looking at a lot of information, a lot of things changing quickly, a lot of uncertainty, it sort of helps you come to a decision when other people can’t. And that’s really the skill. It’s not figuring out the stuff everyone knows or when there’s a lot of data and everything’s clear, it’s when it’s foggy and uncertain, you can still apply the right framework and come to a… relatively confident decision on what to do or not to do. Okay, so if I was sitting down with the CEO, the first thing I’d think of, you know, the three ideas for today, concepts, are jobs to be done, digital superpower number one, and the crazy economics of digital. And the first thing I’d be like is, look, what is the job to be done here? Clearly, Lio Lishuo is doing something different than K through 12 education after school. for people who are trying to do well on the GALCAL and get read better at physics and math and chemistry. Now they’re also doing test prep and they’re also doing some English. But clearly, what they’re attempting to do is not gonna impact the vast majority of New Oriental’s business. And it’s not obvious to me that it can scale out to there, just because you have a good English learning app, how are you gonna teach physics, right? So the first one I’d be like, look, this looks to me like they’re focused on… Helping people, job number one, which I described as an advanced skill, they’re helping people become more proficient in an advanced skill to make a moderate improvement in an advanced skill. That is a very different thing than scoring high on the gaokao or getting in this score on the GMAT. So they have a sort of narrow niche and, you know, what. New Oriental looks a lot more like there’s a lot more going on in the skills area and it’s helping people get into college and get a job and things like that. So I’d say this is mostly a different job to be done. I don’t see it impacting the business at this point outside of English learning at New Oriental. So that’s good. I’d keep an eye on it, but that would be sort of my first, okay, jobs to be done. That’s a useful framework for this scenario because there is actually kind of a lot going on in education. that is not just learning, there’s a lot more going on. Okay, as I look down my little digital superpower list, let’s start at number six, is it scalable? Yeah, it’s really scalable. I mean, this is an app designed to scale from 10 people to a thousand to 10 million to a hundred million to whatever without limit. Now it’s not gonna be global, it’s not Netflix because it’s focused on people who are learning. English who have a Chinese language background, but maybe it could it could go global I suppose. Yeah, it’s highly scalable and New Oriental is not scalable at all. Well, really not at all. Openings, you know, learning center after learning center, not really. Okay, so it’s definitely putting its chips on scalability. Is it viral? I don’t really see any social components or group buying or anything like that, so no, I don’t think it’s terribly viral. Is there a competitive advantage? Are there switching costs? Are there bundling effects? Are there share the consumer mind? Are there low production costs? I wouldn’t call it a competitive advantage. I think obviously what they have is they have an alternate business model that is dramatically cheaper. You can look at competitive advantage a couple of ways. You can say, look, this online teaching company has competitive strengths over an offline teaching company the same way Amazon has competitive strengths in terms of, let’s say, its cost structure versus a traditional retailer, which is true. It’s a different business model. Does it have a competitive advantage against other online businesses? I think this is mostly a business model change is what’s going on. They’re getting rid of a lot of their cost structure. So it doesn’t jump out at me that they have any powerful competitive advantages, but it depends how you define it. Does it have a platform business model? That’s number two. No, there’s no other user group here. It’s consumers. studying online. It’s mainly one user group. Number three, does it have a network effect? This I think is kind of an important question. Network effect is this idea that the more you use something, the more customers you use it, the more that one individual person uses it, it improves in value. It improves the service. So an online merchant marketplace that has 10,000 merchants selling is better. than one that has 100. Does this improve with volume, activity, usage, participation, data, and does that keep improving or does it flatline? That’s an interesting question because people talk a lot about a data network effect. The more you use something, the better the personalization comes. People talk about what they call a learning platform where the more you use something, the smarter it gets. Now they are clearly pitching this as a network effect thing, which is, look, the more people that use this, the smarter the algorithm gets. The more people that use this, we generate more data. The more data we have, the more subtle the algorithm will get. It will pick up smaller accents. It will pick up differences in background and proficiency better and better and better. And that’s at the aggregate level. And at the individual level, the longer you use this as an individual user, the more it will know about you and what you learned and what you haven’t and the service for you will improve more and more. I think there’s a solid chance it has this, but I’m not totally convinced. You know, sometimes when things get better, it’s good for a while, but then it kind of flatlines and it’s easy to replicate. I like Netflix. I like that Netflix personalizes itself to me because it knows what I like to watch and of its whatever thousand shows or whatever, it shows me what I want. I don’t think that’s that hard for another company to replicate very quickly. I think if I go onto Hulu and use it for three weeks, they will begin recommending shows to me approximately as good as Netflix does. So yeah, it’s helpful. I don’t think it lasts that long. I think it’s easy to replicate. So, you know, yeah, it knows me better, but I don’t think it keeps going up. The more and more I use something, I don’t think it gets smarter and smarter. I think after I’ve used Netflix or Hulu for a hundred hours each, I suspect they’re both pretty equivalent in showing me what I wanna see. So that’s the question. Look, if I use another app other than L’Oly Schwo for 10 or 15 lessons, does it begin to give me approximately the same level of lessons as a company like L’Oly Schwo that I’ve been doing for a year? Or is the one I’ve been using for a year dramatically better for me? So I mean, how long does this, if there’s a network effect, a learning advantage, a data network effect, it’s all the same thing. How much does that go on or is it just sort of a, yeah, it’s not that hard to copy. I think that’s a question for them. Spell check is, everyone gives you spell check for free. Once someone figured out how to build a database that could test spelling, it was pretty much a commodity and you could give it away and everybody had it. It wasn’t like the company that invented spell check had some massive network effect. that no one can replicate. What if teaching basic English is more like spell check? That it’s just gonna become a commodity that pretty much anyone can offer. Once you’ve built the database and once you have the algorithm that’s pretty solid, pretty much anyone can teach you English. What if it’s like that? And this is, yeah, they’re first mover, but what they’re creating with the sophisticated AI that can teach you English. What if that just becomes a commodity and pretty much anyone can offer that for nothing? And okay, yeah, maybe Lille Lichua is 5% better, 10% better, but all the copies of it are 80% good enough. Is this just another piece of commodity software? I’m more impressed when people talk about data network effects like this, when it’s something like a mapping service or a traffic service like a Waymo thing, where The value of the app to me is it tells me the traffic in my town right now, because it’s gathering data right now. So the more people using that service right now, this hour, makes it more valuable to me because it’s giving me real time data. And that constantly becomes obsolete and constantly gets refreshed. That to me is a data network effect with real power in it. If it’s just like, hey, here’s some good algorithms for teaching you a basic thing like introduction to English, well, it seems like that’s going to be pretty standard and anyone can offer that. So I like data effects that are more like mapping and things like that as opposed to this. This looks more like spell check to me, but I’m willing to be proved wrong. So I put that as a question mark in my brain. OK, digital superpower number one is where I think the action is here. It’s scalable, number six, and number one, does it dramatically transform the user experience? And clearly that is what they are trying to achieve. And if it does that, I think it actually will be very, very powerful. Like what is it doing that is better than going to New Oriental? Well, number one, it’s gonna be dramatically cheaper. That’s a huge lever to pull because they don’t have people, because they don’t have offices, because they don’t have 700 learning centers. They’ll be able to offer this thing for five to 10 bucks a month, you know, and they’ll be able to offer it for free. Half their classes will be free. You know, freemium model, it’s software we can give it away for free. That’s a huge lever in terms of the consumer experience. You know, forget going down to the local school where they’re gonna charge me $500 per course. I can just watch it on my phone for free. Or I pay $5 a month. That’s a big deal in terms of the consumer experience. Free is a big deal. And that’s why software is one of the reasons I talk so much about the economics of digital is because oftentimes digital lets you do stuff for free. It’s a big deal. Okay, so that would be one. Dramatically improved scalability. talked about convenience, you can watch it anytime, anywhere, you don’t have to go down the street to your store. Personalization, I think personal, like the two that really get my attention in digital education, the digitization of education, are making it dirt cheap, because it’s software and not schools and campuses, and two, making it personalized. Those are the two factors that I really get excited about. I like the idea that if I can get this to work, I will end up sending each of you who are subscribers personalized content based on where I think you are in terms of your understanding, and it’ll be a one-to-one course for everybody, and not one to many. I think that personalization of education is really powerful, and I think they’re trying to do that as opposed to everyone sits in the room and hears the same thing. So the two levers that I think are really important are very cheap and personalization. Second, after that I would say it’s probably convenience. However, it’s not clear that Lio Lishuo, it’s not clear that this works. I think that’s the question mark. You know, does it, do people actually learn well? I mean, is it effective or not? And that’ll be, we’ll just, I just keep watching their adoption numbers. Are people using it? Are more and more people using it every month? Or is this just a matter of a lot of marketing spend to keep roping people in? Are they actually advancing? You know, an AI teacher, which is their core idea, does an AI teacher actually work? And I don’t know yet, but it’s definitely cheaper and it’s definitely more convenient and it’s definitely customizable in a way that regular teaching is not. So that’s where I sort of fall on this one. I think that’s pretty good for today. My takeaway, my advice to the CEO would be, They have a lot of power in scalability. They have a lot of power in leveraging digital economics into dramatically improving the consumer experience if the quality of the education is sufficient. However, this only affects a small portion of your business. And I think the idea of going OMO, online merge offline, I think these companies, particularly New Oriental, they absolutely have to do this. they have to start integrating digital into all of their core products and services. And they can do that internally and they can do that by M&A, which is more or less what they’ve been doing from about 2015. If you look, they’ve been buying companies and investing and moving towards OMO education, where they’re gonna combine their operational footprint across all of these centers with increasing digital tools. But I don’t think they’re gonna jump to this point of, okay, now let’s just get rid of all our teachers. Because it’s not clear to me that works. If I had to place my bet, I would say most of healthcare is gonna be online merge offline, not purely online. But we’ll see. And that’s kind of the lesson for today. The three key concepts for those of you who are subscribers within your various learning goals. Jobs to be done. The old Clayton Quist. Clayton Christensen question, it’s really helpful. I think that’s a really good way to think about education because there’s a lot going on in education like that. That’s concept number one for today. Concept number two is digital superpower number one, which is does it dramatically transform the user experience? I think that is definitely a possibility here. And then concept number three for today is the sexy but dangerous economics of digital. And I think that’s a lot of what’s going on here. I don’t think it’s necessarily a network effect. I don’t think it’s necessarily a platform business model. I think it’s mostly about the fact that when you start to do things with software and not people, the economics become very powerful. You can give things away for free, you’re scalable. It’s just a powerful thing. I think that’s most of what’s going on here is just the economics of digital. So those are the three concepts for today. As for me, I’m doing well. Life in Bangkok is awesome. I was supposed to go back to China pretty quick. I was hoping the restrictions on foreigners entering were about to be lifted, and now we have a new outbreak in Beijing. So I think that plan is dead. My flying back into Beijing is pretty much off the table, I think. So looks like I’m staying here for a little bit longer. Which is fine because things are great here now. Everything’s opened up. I was, for those of you who know Bangkok, I was in Lumpini Park yesterday, which I love Lumpini Park. It’s, you know, it’s the one place all the foreigners know in Bangkok. It’s the nice park, everyone jogs there. And they have the monitors, the monitor lizards. You know, they look like Komodo dragons, but they’re not Komodo dragons. They’re smaller and they sort of swim around in the… the lakes and the rivers of the park. But they’re pretty big. I mean, they like, they sort of crawl around. So I always go down there and I kind of look for the lizards moving around. And it’s funny because they crawl out of the water and they just walk on the paths and all the joggers don’t even think twice. They just run around them. It’s like a total Thailand thing. Like no one even notices, oh, a lizard and they just go around it. And the lizard doesn’t care either. And then it goes back into the water. So I was out there yesterday hunting. not hunting, but looking for lizards, because that’s kind of my thing to do. Hey, there was like seven or eight of them floating around, they’re huge. The kids come out and play with them. It’s great. It’s really a Lumpini Park on a Sunday is really fantastic. So it’s half open now. So I’m going back there and the shopping malls are open. And that’s kind of as far as my life in Bangkok, that’s pretty much all I do. It’s so hot. So you’re either in your apartment, you’re maybe in the park late afternoon or you’re in the shopping center because the AC is on and it’s really important. So that’s kind of the three pillars of my life in Bangkok are back in place as of this week. So it’s great. And I’ll probably bounce down to the islands this next week. I postponed it a little bit. So it looks like now that China’s off the table, I’m heading back down. So I think maybe next week, I’ll be doing this from Koh Tao in. southern Thailand where unfortunately my microphone will not be as good so if the audio quality drops next week, sorry about that. I can’t lug my big blue yeti microphone all the way down there. So anyways, that’s it from me. Best wishes to everybody. Welcome to all the new members. If you’re not a subscriber, please sign up. You can do that at with a 30 day free trial. But welcome to everyone and I hope you’re all doing well and I will talk to you next week.

Leave a Reply